Lecture 9: Unobtrusive Research and the Maine State Legislature

Lesson Overview: Unobtrusive Research

In Chapter 10, Dixon et al. presents the essential case for unobtrusive research through the gathering of already existing data in social science research.  You should recall from an earlier week’s work the Hawthorne Effect, in which the simple act of being observed by researchers leads to a change in the behavior of those being observed.  There are many ways in experiments and surveys to attempt to minimize the effect of being observed by researchers, but what would happen if the researcher’s act of observation could be removed, at least to an indirect distance?  This is the approach of unobtrusive research.

In the lecture below, I invite you to take part in unobtrusive research by gathering existing data on a particular subject (the effect of gender on identity) in a particular domain that matters locally for most of us (the Maine State Legislature).  We will begin with the work underlying every piece of research (theory), and in a DIY activity for the week I challenge you to use the Open Maine Politics website and gather data regarding the gendered work of Maine legislators.

Gender and the Legislature: An Audacious Proposal

A proposal for gender parity in the Canadian parliament, described in the National Post in 2014On September 26, 2014, Canada’s newspaper The National Post described an audacious proposal by former Prime Minister Kim Campbell (Kenez 2014).  Campbell’s idea: to take the Canadian Parliament as it is and divide each seat into two seats: one for a man and one guaranteed to be filled by a woman. Why? Former Prime Minister Campbell says the change will create a more representative system.  “It’s the simplest, least disruptive way,” Campbell explains. “It’s a practice well-known in the Maritimes.  There it was used for the Catholics and Protestants; they used to elect members of each group and it guaranteed equal representation.”

Kim Campbell’s argument can be applied beyond Canada to MaineStateHouseany representative legislature.  In the capital city where UMA is located, we are represented by the Maine State Legislature. Would a division of seats in the Maine State Legislature lead to a better gender representation of Mainers in state government? In order to answer that question, we need to think a little bit more carefully about what representation really is.

This supplemental lecture considers two kinds of political representation related to identities like gender, religion, and race: substantive and descriptive representation.  The video you see below presents this lesson in a visual format.  A text alternative also appears below.


Hanna Pitkin and the Concept of Representation

In 1967, political scientist Hanna Pitkin described two kinds of representation in her book The Concept of Representation (Pitkin 1967). The first of these is descriptive representation. Descriptive representation happens when political leaders resemble the population in identity, in what they look like or some other characteristic held by them. The second kind of representation is substantive representation, when political leaders resemble the population in their policy preferences.  A politician substantively represents those who support Policy X when she or he acts to support Policy X.

To make the distinction clearer, imagine an imaginary circumstance, a simplified legislative body in which we have eight blue members and two red members. Here blues would enjoy a descriptive representation of 80%, since 8 out of 10 members are blue in their identity.

A Congress of Ten, with 8 Blues and 2 Reds

But imagine that in this same hypothetical instance, the two reds and eight blues held different kinds of ideas: pro-red policy and pro-blue policy. As depicted in the example below, pro-red policy enjoys a substative representation of 60%.

A Congress of Ten in which Six support Red interests and Four support Blue interests

DIY Activity #7: Gender and the Maine State Lecture

Now, I’ve described a hypothetical situation above, but we could engage in empirical research to look at real legislative bodies and ask the research question: does descriptive representation leads to substantive representation?  If so, we might generate a research hypothesis that politicians who hold Characteristic X are more likely to support pro-Characteristic-X policy.

If you were going to carry this research out, what would your independent variable be? What would your dependent variable be? How would you operationalize those variables?

Although this is a question about a hypothesis, my question to you isn’t purely hypothetical. To complete this week’s DIY activity, I’d like you to deepen your understanding of substantive representation, descriptive representation, and how the two of them are related to one another by looking at actual data regarding the Maine State Legislature.

Particularly, I’d like you to go to Open Maine Politics at http://openmepolitics.com. This is a freely-available mashup database of Maine state legislative for the current Maine State Legislature. I’d like you to search for some bills on subjects that you believe have to do with the interests of men or women. I want you to find at least two bills that in your judgment, if passed, would lead to better substantive representation for either men or women Mainers.

An arrow highlighting the search box at Open Maine Politics

Once you have made a successful search and identified at least two bills that you think clearly are related to the interests of some identity group, I want you to take a close look at the sponsors for those bills. Find out what the gender identity each of the sponsors is, and then compare that to the gender identity of status of ALL senators and representatives in the entire Maine State Legislature. Now compare the two: the status of those who support the bill that you believe promotes the interest of a particular identity’s interest, and the identity of legislators altogether. That comparison will give you an answer to our research question.

In order to get the answer to that research question, you’ll have to frame your hypothesis, you’ll have to operationally define an independent variable and a dependent variable, and you’ll have to gather data.

Once you’ve completed your work, don’t keep your findings all to yourself!  Report all your work, including your variables, hypotheses, data and your results, in a word processing document. Upload your work to the area under the “DIY Activities” link of our course Blackboard page called “DIY Activity #7: Gender and the Maine State Legislature.”


Kenez, Hayden. 2014. “Two Paths to Gender Parity in the House.” The National Post September 26: A1.

Pitkin, Hanna Fenichel. 1967. The Concept of Representation. Berkeley: University of California Press.

Leave a Reply

Your email address will not be published. Required fields are marked *